Green roofs have been proven to help reduce Solar panel's heat to increase its performance. However, finding the type of plants that provides the best results for a specific climate still a challenging task that we will try to solve in this article.
A green roof, or rooftop garden, is a vegetative layer grown on a rooftop. They provide shade, remove heat from the air, and reduce temperatures of the roof surface and surrounding air (which includes the solar panels).
Using green roofs in cities or other built environments with limited vegetation can moderate the heat effect, particularly during the day. Green roof temperatures can be 10–30°C lower than those of conventional roofs and can reduce city-wide ambient temperatures by up to 7°C. In addition, green roofs can reduce building energy use by 0.7% compared to conventional roofs, reducing peak electricity demand and leading to an important annual savings.
Our solution :
To find the most appropriate type of plants, we've created an embedded device that can retrieve temperature, humidity and RGB components of the panel's nearby environment.
It was designed to be a standalone device that grabs its energy from a tiny integrated solar panel. An extra software layer and some hardware manipulations were added to enhance its performance and autonomy(source code on the next section).
As suggested by our customer (food-industry section of Sorbonne University), the device must grab the previous physical parameters from both under and near the solar panel. These two parameters will simplify the study of selecting the optimal position of the device also.
The data collected from our device are transmitted to our database using SIGFOX network. SIGFOX utilizes a wide-reaching signal that passes freely through solid objects, called "Ultra Narrowband" and requires little energy, being termed "Low-power Wide-area network (LPWAN)".
We have also designed an online Graphical User Interface that helps analyzing the data and that can generate reports as charts or csv files.
I - How to reproduce our solution ?
I.1 - Schematics
Prepare all the components joined in the download section of this article. Using a breadboard and some switches, make the following connections of the sensors to the microcontroller.
I.2 - PCB
I.3 - µC Programming
To program our microcontroller, we've used MBED. You can download the code joined in this article. Due to some problems that we encountered during the development phase, there many work-around that we've made to make all the libraries work together on the NUCLEO-L432KC. You can download the patched libraries from our github's project directory.
I.4 - SIGFOX backend:
We have used this tutorial to configure our sigfox backend with our website. After formatting the data properly, you need to point it to the public IP address of the virtual machine.
I.5 - Website :
As students, we have used Microsoft Azure free credits (100$) to create a virtual machine. We have installed on it a LAMP server to hold our website. This virtual machine will receive the SIGFOX callbacks using its public IP address. You can find the source code in the same github directory.
II- User Manual :
- Fully charge the battery with the power modem.
- Connect the battery and the solar panel to the power module.
- Your module is powered and is working.
- Put the main module next to your solar panel and the secondary module below your solar panel, which will allow you to evaluate the impact of plants with and without the solar panel.
- Login to account in website with sigfox ID.
- You can go to the website to see the data every 15 minutes or directly to the test site by pressing the button to view the data on the LCD.
- Website : https://panneautech.000webhostapp.com/ or using the VM public IP http://18.104.22.168/
- Username : polytech
- Password : sorbonne
- You have possibility to extract web data with.CSV and open in Excel or text editor.